The proposed studies will examine regional brain dysfunction in schizophrenia by measuring local cerebral blood flow (CBF) in patients and normal controls during performance of verbal and facial memory tasks. The focus on learning and memory is prompted by evidence that these functions are disproportionately impaired in schizophrenia. The studies proposed in this RO1 will use the "Neurobehavioral Probe" method to help identify regions of abnormal activation during performance of memory tasks. CBF will be measured with PET using 15-Oxygen labelled water (the "slow bolus" method). Subjects will be patients with schizophrenia and normal controls (balanced socio-demographically), and they will each be studied at resting baseline and 3 activation conditions. The experimental dimensions to be examined are learning (acquisition) vs recall and verbal vs facial stimuli. There will be 60 subjects per cell, 30 men 30 women, 15 patients with schizophrenia and 15 normal controls per gender. Based on results of these experiments and neuropsychological studies performed in the MHCRC, other dimensions which show differential impairment in schizophrenia can be examined. These could be list vs paired-associate learning, emotional vs nonemotional, or "procedural" vs "declarative" memory. The hypothesis for normal controls is that the experimental conditions will differentially activate brain regions engaged in learning and retrieval processes. There is some convergence of evidence for the primary role in memory of the hippocampus and associated cortex, temporal and dorsolateral frontal cortical regions. In normals we can test the hypothesis that activation will be greater for acquisition relative to recall, with laterality effects for words (L>R) vs faces (R>L). In earlier studies of cortical CBF we found specificity of activation to be associated with better performance, and this can be extended in the present investigation to subcortical regions. We will examine which aspects of this system are impaired in schizophrenia, and test the hypothesis that abnormalities in activation are related to symptom severity and neuropsychological deficits.